> Rational investors have no problem whatsoever projecting the impact to future cash flows and adjusting the amount they're willing to pay now.

That's not accounting for the time value of money.

Suppose you have a rental unit and before rent control you were planning to rent it out. That now has far less attractive returns and it suddenly makes more sense to sell it as a condo to an occupant rather than keeping it to rent it out. Likewise, other prospective landlords no longer want to buy it at the previous market price (returns went down and they can invest in stocks or housing in some other city instead), so the short-term effect is to increase the number of property sellers and decrease the number of buyers. Short-term, property values going down is the expected thing.

But construction going down is also the expected thing, for the same reason. If property values are lower then the number of viable construction projects is lower and less construction happens.

The lack of construction then continues until rents, even after rent control, are high enough to justify more construction, i.e. are even higher than they were originally, because now to justify the same investment as before you need the current rent to be high enough to account for the inability to increase it later. And with less construction happening, that's the natural result in a growing area. More people have to bid on the same number of units, rents go up. So the short-term effect is lower real estate prices, the long-term effect is higher.

Now you say, if we expect real estate costs to be higher later, why don't investors take advantage? Which is the "time value of money" issue. If you invest there now because the prices will be high later, what do you do with the property in the meantime? If you don't rent it out, paying $1 today to get $1 in ten or twenty years is dumb, so you only buy if the current price is at a discount. If you do rent it out, then you'd be stuck trying to sell a building with a rent-controlled tenant, which isn't worth as much as the same building as empty as you bought it which you could sell as condos or rent out at current rents instead of price controlled ones, and then you still need a discount. And so the current seller has to provide a discount even if the real estate costs will be higher in a few years.

This argument presupposes multiple levels of assumption. By the point where the assertion is made that construction costs would drop, the prediction's error bars equal the entire range.

There's no reason to believe that construction companies would accept jobs at prices below the cost of materials and labor. Construction companies frequently let workers go rather than accept large negative cash flows.